Phenotyping effort

Collaborative phenotyping effort of European Drosophila melanogaster populations

Genomic analyses (Kapun et al. 2020) showed that there is longitudinal population structure, continent-wide sweeps, candidate genes for local climate adaptation by using 48 pooled population samples from 32 locations.

In 2017, DrosEU consortium has decided to make a collaborative phenotyping effort of European Drosophila melanogaster populations at the Groningen meeting.

The call was made with approximate numbers of populations, lines, etc. and under predefined criteria (finish within a year, reps, blocks, etc) and participating labs have indicated which traits they are willing to phenotype.

Samples

Nine sampling locations were chosen based on genomic data and cover a wide range of latitude (~20°) and longitude (~40°) across the continent (see table below). From each location, 15 to 20 isofemale lines were established in corresponding labs at the sampling location and the isofemale lines were centrally maintained by Élio Sucena at Instituto Gulbenkian de Ciência (IGC), Lisbon, Portugal. A total of 173 isofemale lines were used in this study.

##    Country           Location Latitude Longitude Altitude            Collectors
## 1 Portugal            Recarei   41.150    -8.410      175          Jorge Vieira
## 2    Spain Gimenells (Lleida)   41.618     0.620      173       Josefa Gonzales
## 3  Denmark        Karensminde   55.945    10.213       15            Mads Schou
## 4  Germany             Munich   48.180    11.610      520 Amanda Glaser-Schmitt
## 5  Austria        Mauternbach   48.375    15.560      572     Andrea Betancourt
## 6  Finland               Akaa   61.100    23.520       88        Maaria Kankare
## 7  Ukraine               Uman   48.753    30.206      214      Iryna Kozeretska
## 8   Turkey            Yesiloz   40.231    32.260      680            Banu Onder
## 9   Russia             Valday   57.979    33.244      217       Elena Pasyukova

## Scale on map varies by more than 10%, scale bar may be inaccurate

Contributed labs are Abbott, Bergland, Billeter, Colinet, Flatt, Fricke, Gibert, Gonzalez, Grath, Hoedjes, Kozeretska, Onder, Parsch, Pasyukova, Posnien, Ritchie, Schlötterer, Schmidt, Stamenkovic-Radak, Tauber, Vieira, Wegener, Zwaan, Mensch and detailed list below :

droseu_labs

##            Country             Lab                PI                     Trait
## 1           Sweden            Lund            Abbott              Pigmentation
## 2              USA Charlottesville          Bergland                  Diapause
## 3  The Netherlands       Groningen          Billeter                 Fecundity
## 4           France          Rennes           Colinet                Dry weight
## 5      Switzerland        Fribourg             Flatt                  Diapause
## 6      Switzerland        Fribourg             Flatt                  Lifespan
## 7          Germany        Muenster            Fricke                 Fecundity
## 8           France            Lyon            Gibert          Development time
## 9           France            Lyon            Gibert              Pigmentation
## 10          France            Lyon            Gibert                 Viability
## 11           Spain       Barcelona          Gonzalez      Cold-shock mortality
## 12           Spain       Barcelona          Gonzalez     Starvation resistance
## 13         Germany          Munich             Grath          Development time
## 14         Germany          Munich             Grath                 Viability
## 15 The Netherlands        Lausanne           Hoedjes          Development time
## 16 The Netherlands        Lausanne           Hoedjes                Dry weight
## 17 The Netherlands        Lausanne           Hoedjes                 Viability
## 18         Ukraine            Kyiv        Kozeretska      Cold-shock mortality
## 19         Ukraine            Kyiv        Kozeretska             Thorax length
## 20          Turkey          Ankara             Onder                Dry weight
## 21          Turkey          Ankara             Onder     Starvation resistance
## 22          Turkey          Ankara             Onder                 Wing area
## 23         Germany          Munich            Parsch      Heat-shock mortality
## 24         Germany          Munich            Parsch                  Lifespan
## 25          Russia          Moscow         Pasyukova                  Lifespan
## 26          Russia          Moscow         Pasyukova     Starvation resistance
## 27         Germany       Göttingen           Posnien             Thorax length
## 28         Germany       Göttingen           Posnien                 Wing area
## 29              UK      St Andrews           Ritchie             Thorax length
## 30              UK      St Andrews           Ritchie                 Wing area
## 31              UK      St Andrews           Ritchie          Wing patterning 
## 32         Austria          Vienna       Schlötterer                  Diapause
## 33             USA    Philadelphia           Schmidt          Development time
## 34             USA    Philadelphia           Schmidt              Pigmentation
## 35             USA    Philadelphia           Schmidt             Thorax length
## 36             USA    Philadelphia           Schmidt          Time to pupation
## 37             USA    Philadelphia           Schmidt                 Viability
## 38          Serbia        Belgrade Stamenkovic-Radak          Development time
## 39          Serbia        Belgrade Stamenkovic-Radak                 Viability
## 40          Serbia        Belgrade Stamenkovic-Radak                 Wing area
## 41          Israel           Haifa            Tauber        Locomotor activity
## 42        Portugal           Porto            Vieira  Chill-coma recovery time
## 43        Portugal           Porto            Vieira      Cold-shock mortality
## 44        Portugal           Porto            Vieira      Heat-shock mortality
## 45         Germany        Würzburg           Wegener Circadian eclosion timing
## 46 The Netherlands      Wageningen             Zwaan          Development time
## 47 The Netherlands      Wageningen             Zwaan                 Viability
## 48       Argentina    Buenos Aiers            Mensch  Chill-coma recovery time

FIX CITIES COORDINATES!!!! ST ANDREWS IS IN THE SEA!!!!

Phenotyping experiments

Overview

Methods

Phenotyping

Contributors and their experimental methods

Viability

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Contributors : Patricia Gibert, Sonja Grath, Paul Schmidt, Marina Stamenkovic-Radak, Bas Zwann, Joost van den Heuvel, Vieira Cristina, Mouton Laurence, Kremer Natacha, Martinez Sonia, Ingo Müller, Hristina Kostic, Amy Goldfischer, Paul Schmidt, Ozan Kiratli, Yonatan Babore, Liam Forsythe, Marija Savic Veselinovic, Marija Tanaskovic, Aleksandra Patenkovic, Mihailo Jelic, Katarina Eric, Pavle Eric, Slobodan Davidovic

3-5 day old adults (at least 25 pairs) are allowed to lay eggs en masse. Yeast is provided to stimulate egg laying (for at least 2 hours). Eggs are collected, and 40 are placed in each vial. Viability is calculated per vial, as the percentage of individuals that emerged from the 40 eggs.

Please see below number of

Add methods for Schmidt Lab Replicate vials

Developmental time

Egg-to-pupa developmental time

Schmidt Lab

Paul Schmidt, ….

Egg-to-adult developmental time

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Patricia Gibert, Katja Hoedjes, Paul Schmidt, Marina Stamenkovic-Radak, Bas Zwann, …. Developmental time is scored as both the egg-to-pupa and egg-to-adult development time. Both were scored twice a day, when the chamber lights are turned on and two hours before they are turned off. In order to measure the egg-to-pupa developmental time, the spot where a pupa is found is marked with a permanent marker to keep track of which pupae have emerged in each day. The egg-to-adult developmental time is estimated by counting all emerged adults from the vial, and by sexing them.

Dry weight

Colinet, Hoedjes and Onder Labs

Herve Colinet, Katja Hoedjes, Banu Onder, ….

At day 7 after emergence, flies are killed by snap freezing them in liquid nitrogen, by putting them at -20ºC or by putting them into an ethyl acetate solution and stored at -20ºC. Then they are sexed and placed into 96 wells plates, and placed in an oven set at 60-70 °C, for at least 3 days. At this point flies can be stored at room temperature using a protective cover. If this is the case, the day before measurements are made dry flies are again placed for 24h in the oven (60-70 °C) to ensure material is well dehydrated. Flies are then placed on a small piece of aluminium foil for direct weight measurement on microbalance (accuracy 1µg).

Thorax length

Kozeretska, Posnien, Ritchie and Schmidt Labs

Five to seven days old flies are placed onto a double-sided sticky tape attached to a microscope slide and a picture of the thorax taken using a digital camera connected to a dissecting microscope. The same magnification and resolution is always used to increase reproducibility, and a scale bar inserted on each photo to allow transforming pixels into µm units. Thorax length is defined as the distance from the anterior margin of the thorax to the posterior tip of the scutellum and it is measured using the “Straight Line” in ImageJ/Fiji.

Wing area

Onder, Posnien, Ritchie and Stamenkovic-Radak Labs

Both the left and right wings of five to seven days old flies (10 flies per sex per replicate) are removed and placed into a drop of Entellan®Merck, Hoyer’s Medium, sticked to a double-side sticky tape, or taped directly to the slide. Pictures of the wing preparations are taken using a digital camera connected to a dissecting microscope. The same magnification and resolution is always used to increase reproducibility. A scale bar is placed on each photo to allow transforming pixels into µm units. Manual measurements of wing length and wing area are performed using the “Straight Line” and “Polygon Selection” tools, respectively of ImageJ/Fiji (10.1371/journal.pone.0000007).

Fecundity

Billeter and Fricke Labs

For each isofemale line, 10 males and 10 females are placed together in single-sex groups and allowed to mature for five days. Then, they are placed together (5-7 pairs), and mating interactions observed to ensure successful mating (at least 10 min copulation duration) to ensure that we have five successfully mated females. After a successful mating, males are discarded and females allowed to oviposit alone for 48 hours, moved to another vial, and allowed to oviposit for four days, and again moved to another vial and allowed to oviposit for two days to check that of egg-laying stopped. Vials are incubated until all offspring is born. Individuals are then frozen and the offspring counted.

Lifespan

Flatt, Parsch and Pasyukova Labs

Esra Durmaz, Envel Kerdaffrec, TBO, Virginie, John Parsch, Amanda, Eliza, Lena Pasyukova, Alexander

Line level lifespan: Ten flies per sex/line are placed in each vial. The age at death will be scored when changing the food, at least three times a week. Five replicates are used.

Population level lifespan: Flies are kept in 1L demography cages (5 flies per line/sex for each population). The age at death will be scored when changing the food, at least three times a week. Ten replicates were performed.

Cold-shock mortality

Gonzalez, Kozeretska and Vieira Labs

Batches of 15-20 seven days old flies are placed for 18 hours in an empty vial immersed in an ice-water slurry box placed at a 4°C room for 18 hours. Then the vials are removed to a bench in a 25°C room and mortality scored 24 hours later.

Chill-coma recovery time

Vieira and Mensh Labs

Sexed flies are placed in an empty vial immersed in an ice-water slurry box placed at a 4°C room in the morning. Six hours later, flies are removed from the tube to individual wells of 24 well plates while being kept on ice. A timer is started once the plate is moved from the ice to a bench in a 25°C room. Each fly is checked by eye for recovery for a maximum of 60 minutes. Flies that are able to stand on their legs are considered recovered and the CCRT (in seconds) recorded.

Heat-shock mortality

Parsch and Vieira Labs

Batches of 15-20 seven days old flies are placed in empty vials inside a 37ºC incubator and mortality checked for 7 hours every 30 minutes.

Diapause

Bergland, Flatt and Schlotterer Labs

In order to induce diapause, two hour old ‘phenotypic virgins’ (pharate or melanized with meconium visible) female flies are exposed to 12°C and 10:14 light/dark hours for 3 weeks, using an incubator that allows temperature tracking. Vials are changed once per week. After three weeks under diapause conditions, flies are frozen at -80C until dissection. Both ovaries will be examined and classified according to the following simplified ‘classification’: 1) < stage 10: diapause; 2) stage 10-13: intermediate; 3) stage 14: non-diapause.

Standard diapause conditions: 12°C at 10:14 hrs L:D (under these conditions we might expect around 50% dipause frequency)
duration of exposure: 3 weeks
assays will be done in incubators (Percival or Memmert); temperature conditions will be tracked using data loggers / iButtons
Try to have haphazard standardization; similar conditions coupled with consistent tracking of temperature (multiple iButtons, for example)
We will use egg lays in bottles to be able to collect more flies for the assay (approx. 1-2 bottles per line)
Larval density will be controlled (approximately) prior to assays: we will do this by letting a fixed number of (approx. 10-20) females (housed together with 5-10 males) oviposit over night (fixed time window) - how many females and the duration of egg laying might need to be adjusted a bit depending on the fertility of the lines (to be determined when lines are being amplified prior to assays)
The Bergland and Flatt labs will standardize their fly food according to the Gulbenkian recipe (in terms of amounts / proportions); Manolis from the Schlötterer lab will not standardize the food and use their local diet; importantly, flies for the assays will not be exposed to any (additional) live yeast (i.e., the diet will contain liophilized yeast but no live yeast sprinkled on top).
From the bottles we will collect upon eclosion ‘phenotypic virgins’ (=pharate adults or melanized with meconium visible), approx. within 2 hours after eclosion
Flies will be kept in vials for assay (15-20 flies in the same vial to account for mortality); we aim to measure a minimum of 10 females per line (from approx. 10-15 lines x 10 populations, yielding about 1000-1500 flies for dissections in total)
The food in the assays will be changed once per week
To keep the workload and effort manageable the assays will be performed in 2-3 blocks (optimally only 2) over time: e.g., from each population we will measure a subset of all lines (e.g., 5-7 lines, i.e. 1/2 or 1/3 of all lines) in each of the 2-3 blocks.
After 3 weeks under diapause conditions, flies will either be dissected or kept frozen until dissection (dry, at -80C; can be stored for a few months)
We will adopt a simplified diapause ‘classification’:
Stage 10 will be defined as the checkpoint
< Stage 10: diapause
Stage 10-13: intermediate
Stage 14: non-diapause
We will examine both ovaries.

Circadian eclosion timing

Wegener Lab

Locomotor activity

Tauber Lab

The locomotor activity of the flies (males only) was measured using the DAM2 Drosophila monitors (Trikinetics Inc., Waltham, MA). Flies were 1-3 days old. Single flies were placed in glass tubes (10 cm × 0.5 cm) that were filled with 2 cm sugar/agar medium. The monitors were placed in light chambers driven by LED, in an incubator at 24°C, ~30% humidity. The flies were entrained to a light-dark cycle (LD 12:12) for 5d and then allowed to free-run for 10 d in constant darkness (DD). The activity data were processed into 30 min bins, and four different variables were analysed. These included the circadian period, and the phase, which were analyzed using the FFT NLLS algorithm available at the BioDare2 server (https://biodare2.ed.ac.uk/). The other two variables, level of activity and the nocturnal/diurnal ratio were analyzed by a custom-made R script.

Starvation resistance

Gonzalez, Onder and Pasyukova Labs

Batches of 10 sexed flies are transferred to glass vials filled with 5 mL of 2% agar for starvation 3-7 days after eclosion. The age at death will be scored every 8 hours.

Pigmentation

Abbott, Gibert and Schmidt

For each line, 10 females, 13-15 days old, either alive or stored in 95% ethanol, are air dried and placed on their left side, and pictures taken using a dissecting microscope. Images are then analysed in ImageJ 1.46r, using the Area Fraction measurement in the Analyze menu. Area Fraction measures the percentage of pixels in a selected area that have been highlighted in red using the Threshold tool. This gives an estimate of the percentage of dark pigmentation on the three terminal tergites of the abdomen (4, 5 and 6).

Data reformatting

MasterSheets

Insert a MasterSheet example

Data check

For each trait, sex and lab, we run counts for population, line and replicate vial (if applicable).

Results I : Linear Models

setwd(localgit)

Viability

Contributors: Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Data reformatting

Read data in R

d_Via <- read.csv("MasterSheets_Oct21_git/VIA_MasterSheet_Oct21.csv")
str(d_Via)

## 'data.frame':    2367 obs. of  12 variables:
##  $ Supervisor.PI               : chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet                        : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                       : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : chr  "AK1" "AK1" "AK1" "AK10" ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : chr  "Gibert_1_AK1_1" "Gibert_1_AK1_2" "Gibert_1_AK1_3" "Gibert_1_AK10_1" ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : int  88 88 88 88 88 88 88 88 88 88 ...

Factors need reformatting (i.e. Supervisor.PI should be coded as a factor, not character).

str(d_Via)

## 'data.frame':    2367 obs. of  12 variables:
##  $ Supervisor.PI               : chr  "Gibert" "Gibert" "Gibert" "Gibert" ...
##  $ Diet                        : chr  "NS" "NS" "NS" "NS" ...
##  $ Batch                       : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : chr  "AK1" "AK1" "AK1" "AK10" ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : chr  "Gibert_1_AK1_1" "Gibert_1_AK1_2" "Gibert_1_AK1_3" "Gibert_1_AK10_1" ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : chr  "Finland" "Finland" "Finland" "Finland" ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : int  88 88 88 88 88 88 88 88 88 88 ...

d_Via$Supervisor.PI <- as.factor(d_Via$Supervisor.PI)
d_Via$Diet <- as.factor(d_Via$Diet)
d_Via$Batch <- as.factor(d_Via$Batch)
d_Via$Population_Lat <- factor(d_Via$Population, levels= c("YE","RE","GI","MU","MA","UM","KA","VA","AK"))
d_Via$Population_Lon <- factor(d_Via$Population, levels= c("RE","GI","KA","MU","MA","AK","UM","YE","VA"))
d_Via$Population_Alt <- factor(d_Via$Population, levels= c("KA","AK","GI","RE","UM","VA","MU","MA","YE"))
d_Via$Line <- as.factor(d_Via$Line)
d_Via$ReplicateVial <- as.factor(d_Via$ReplicateVial)
d_Via$ProportionEggtoAdultSurvival <- as.numeric(d_Via$ProportionEggtoAdultSurvival)
d_Via$Country <- as.factor(d_Via$Country)
d_Via$Latitude <- as.numeric(d_Via$Latitude)
d_Via$Longitude <- as.numeric(d_Via$Longitude)
d_Via$Altitude <- as.numeric(d_Via$Altitude)

Now they should be in the correct format, see below.

str(d_Via)

## 'data.frame':    2367 obs. of  15 variables:
##  $ Supervisor.PI               : Factor w/ 6 levels "Gibert","Grath",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Diet                        : Factor w/ 1 level "NS": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Batch                       : Factor w/ 4 levels "1","2","3","4": 1 1 1 1 1 1 1 1 1 1 ...
##  $ Population                  : chr  "AK" "AK" "AK" "AK" ...
##  $ Line                        : Factor w/ 172 levels "AK1","AK10","AK11",..: 1 1 1 2 2 2 4 4 4 9 ...
##  $ ReplicateVialOld            : int  1 2 3 1 2 3 1 2 3 1 ...
##  $ ReplicateVial               : Factor w/ 2367 levels "Gibert_1_AK1_1",..: 1 2 3 4 5 6 7 8 9 10 ...
##  $ ProportionEggtoAdultSurvival: num  0.68 0.73 0.63 0.85 0.75 0.8 0.85 0.88 0.7 0.68 ...
##  $ Country                     : Factor w/ 9 levels "Austria","Denmark",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ Latitude                    : num  61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 61.1 ...
##  $ Longitude                   : num  23.5 23.5 23.5 23.5 23.5 ...
##  $ Altitude                    : num  88 88 88 88 88 88 88 88 88 88 ...
##  $ Population_Lat              : Factor w/ 9 levels "YE","RE","GI",..: 9 9 9 9 9 9 9 9 9 9 ...
##  $ Population_Lon              : Factor w/ 9 levels "RE","GI","KA",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ Population_Alt              : Factor w/ 9 levels "KA","AK","GI",..: 2 2 2 2 2 2 2 2 2 2 ...

Descriptive Statistics

Tables for descriptive statistics at population and line level can be found in the main directory, under the file name table_TraitAbbreviation_Level_BatchInfo.csv (i.e. table_Via_Line_wobatch.csv)

Descriptive statistics for viability at the line level, with batch information :

Descriptive statistics for viability at the line level, without batch information :

Descriptive statistics for viability at the population level, with batch information :

Descriptive statistics for viability at the population level, without batch information :

Plots and Linear Models by Lab

The very same theme has been used all across the document, being ggplot theme

droseu_theme <- theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black",),  axis.title.x = element_text(size = 16), axis.text.x = element_text(size = 16),axis.text.y = element_text(size = 16),axis.title.y = element_text(size = 16))

Data range is calculated with #r min() and #r #max() functions for each trait

min_Via <- min(d_Via$ProportionEggtoAdultSurvival)
max_Via <- max(d_Via$ProportionEggtoAdultSurvival)

y-axis is scaled by the minimum (0) and maximum (1) values in the full data set.

Here is an example code for egg-to-adult viability, Schmidt Lab. The same code is used to generate plots for other contributing labs by filtering the data at supervisor level (for females and males, if applicable).

Schmidt Lab

pdf(file="p_Via_Schmidt.pdf",width=8, height=5)
p_Via_Schmidt <- ggplot(data = (subset(d_Via,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=ProportionEggtoAdultSurvival, fill=Batch)) + geom_boxplot(outlier.shape = NA, notch=FALSE) +  geom_jitter(width=0.2,alpha=0.05) + labs(title="p_Via_Schmidt", x="Population", y = "ProportionEggtoAdultSurvival")

p_Via_Schmidt + coord_cartesian(ylim = c(min_Via, max_Via))+ droseu_theme

invisible(dev.off())

pdf(file="p_Via_pop_Schmidt.pdf",width=8, height=5)
p_Via_pop_Schmidt <- ggplot(data = (subset(d_Via,Supervisor.PI=='Schmidt')), aes(x=Population_Lat, y=ProportionEggtoAdultSurvival, fill=Population)) +geom_boxplot(outlier.shape = NA, notch=FALSE) +  geom_jitter(width=0.2,alpha=0.05) + labs(title="p_Via_pop_Schmidt", x="Population", y = "ProportionEggtoAdultSurvival") 
dev.off()

## quartz_off_screen 
##                 2

p_Via_pop_Schmidt + coord_cartesian(ylim = c(min_Via, max_Via))+ droseu_theme

invisible(dev.off())

Via_lmer_Schmidt <- lm(ProportionEggtoAdultSurvival ~ Population, data = d_Via[d_Via$Supervisor.PI == "Schmidt",])
capture.output(summary(Via_lmer_Schmidt),file = "Via_lmer_Schmidt_sum.txt")
capture.output(anova(Via_lmer_Schmidt),file = "Via_lmer_Schmidt.txt")
capture.output(emmeans(Via_lmer_Schmidt, list(pairwise ~ Population), adjust = "tukey"),file = "Via_lmer_Schmidt_tk.txt")

anova(Via_lmer_Schmidt) #lm()` is used only for Schmidt Lab's viability data, as only one vial per line was phenotyped

## Analysis of Variance Table
## 
## Response: ProportionEggtoAdultSurvival
##             Df Sum Sq  Mean Sq F value  Pr(>F)  
## Population   8 1.0046 0.125579  2.5425 0.01256 *
## Residuals  153 7.5570 0.049392                  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Schmidt)

## 
## Call:
## lm(formula = ProportionEggtoAdultSurvival ~ Population, data = d_Via[d_Via$Supervisor.PI == 
##     "Schmidt", ])
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.48129 -0.19237  0.02616  0.16173  0.51871 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   0.625157   0.049695  12.580   <2e-16 ***
## PopulationGI -0.143866   0.075911  -1.895   0.0600 .  
## PopulationKA  0.135186   0.070280   1.924   0.0563 .  
## PopulationMA  0.007477   0.072206   0.104   0.9177    
## PopulationMU -0.052544   0.070280  -0.748   0.4558    
## PopulationRE  0.004666   0.075911   0.061   0.9511    
## PopulationUM  0.085505   0.077444   1.104   0.2713    
## PopulationVA -0.041436   0.070280  -0.590   0.5563    
## PopulationYE -0.087422   0.070280  -1.244   0.2154    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.2222 on 153 degrees of freedom
## Multiple R-squared:  0.1173, Adjusted R-squared:  0.07119 
## F-statistic: 2.542 on 8 and 153 DF,  p-value: 0.01256

Stamenkovic-Radak Lab

anova(Via_lmer_StamenkovicRadak)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 0.39276 0.049095     8 155.26  5.0862 1.243e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_StamenkovicRadak)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "StamenkovicRadak", ]
## 
## REML criterion at convergence: -589.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.6754 -0.5185  0.0052  0.5759  2.2517 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.013176 0.11479 
##  Batch           (Intercept) 0.001207 0.03475 
##  Residual                    0.009652 0.09825 
## Number of obs: 501, groups:  Line:Population, 167; Batch, 4
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## (Intercept)    0.67693    0.03355  21.98005  20.176 1.13e-15 ***
## PopulationGI  -0.07436    0.04570 155.38082  -1.627  0.10573    
## PopulationKA  -0.02793    0.04054 155.23842  -0.689  0.49191    
## PopulationMA  -0.12034    0.04051 155.10198  -2.971  0.00344 ** 
## PopulationMU   0.02403    0.04051 155.08224   0.593  0.55386    
## PopulationRE  -0.05090    0.04302 155.34042  -1.183  0.23859    
## PopulationUM  -0.08775    0.04168 155.34524  -2.106  0.03685 *  
## PopulationVA  -0.11615    0.04051 155.10172  -2.867  0.00472 ** 
## PopulationYE  -0.18226    0.04054 155.23842  -4.496 1.35e-05 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI -0.535                                                 
## PopulatinKA -0.606  0.443                                          
## PopulatinMA -0.603  0.441  0.499                                   
## PopulatinMU -0.605  0.442  0.500  0.500                            
## PopulatinRE -0.571  0.415  0.472  0.471  0.472                     
## PopulatinUM -0.590  0.429  0.488  0.486  0.487  0.460              
## PopulatinVA -0.604  0.445  0.500  0.499  0.500  0.470  0.485       
## PopulatinYE -0.606  0.443  0.501  0.499  0.500  0.472  0.488  0.500

Zwaan Lab

anova(Via_lmer_Zwaan)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 1.1594 0.14492     8 151.78  6.6463 1.944e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Zwaan)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Zwaan", ]
## 
## REML criterion at convergence: -328.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.8665 -0.4405  0.1045  0.5674  2.7223 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.01052  0.1025  
##  Batch           (Intercept) 0.00000  0.0000  
##  Residual                    0.02180  0.1477  
## Number of obs: 524, groups:  Line:Population, 169; Batch, 2
## 
## Fixed effects:
##                Estimate Std. Error         df t value Pr(>|t|)    
## (Intercept)    0.753292   0.029612 148.129254  25.439  < 2e-16 ***
## PopulationGI  -0.109677   0.045264 147.538931  -2.423   0.0166 *  
## PopulationKA   0.006964   0.042022 149.533027   0.166   0.8686    
## PopulationMA  -0.039616   0.042055 150.208886  -0.942   0.3477    
## PopulationMU  -0.045938   0.042393 153.358912  -1.084   0.2802    
## PopulationRE  -0.107348   0.044247 145.966515  -2.426   0.0165 *  
## PopulationUM   0.025615   0.043113 148.966551   0.594   0.5533    
## PopulationVA  -0.020760   0.042344 153.060112  -0.490   0.6247    
## PopulationYE  -0.226033   0.042298 153.374414  -5.344 3.24e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI -0.654                                                 
## PopulatinKA -0.705  0.461                                          
## PopulatinMA -0.704  0.461  0.496                                   
## PopulatinMU -0.698  0.457  0.492  0.492                            
## PopulatinRE -0.669  0.438  0.472  0.471  0.467                     
## PopulatinUM -0.687  0.449  0.484  0.484  0.480  0.460              
## PopulatinVA -0.699  0.457  0.493  0.492  0.488  0.468  0.480       
## PopulatinYE -0.700  0.458  0.493  0.493  0.489  0.469  0.481  0.490
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

Hoedjes Lab

anova(Via_lmer_Hoedjes)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq  Mean Sq NumDF DenDF F value    Pr(>F)    
## Population 0.35451 0.044314     8   158  6.1992 5.871e-07 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Hoedjes)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Hoedjes", ]
## 
## REML criterion at convergence: -712.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.7848 -0.5126 -0.0026  0.5498  2.7446 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev. 
##  Line:Population (Intercept) 1.219e-02 1.104e-01
##  Batch           (Intercept) 2.196e-12 1.482e-06
##  Residual                    7.148e-03 8.455e-02
## Number of obs: 501, groups:  Line:Population, 167; Batch, 4
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## (Intercept)    0.77450    0.02699 157.99983  28.694  < 2e-16 ***
## PopulationGI  -0.17183    0.04123 157.99984  -4.168 5.06e-05 ***
## PopulationKA  -0.05633    0.03817 157.99984  -1.476 0.141995    
## PopulationMA  -0.06600    0.03817 157.99984  -1.729 0.085761 .  
## PopulationMU  -0.03500    0.03817 157.99984  -0.917 0.360593    
## PopulationRE  -0.15117    0.04123 157.99985  -3.666 0.000336 ***
## PopulationUM  -0.08960    0.03982 157.99985  -2.250 0.025827 *  
## PopulationVA  -0.09367    0.03817 157.99984  -2.454 0.015222 *  
## PopulationYE  -0.21483    0.03817 157.99984  -5.628 8.11e-08 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI -0.655                                                 
## PopulatinKA -0.707  0.463                                          
## PopulatinMA -0.707  0.463  0.500                                   
## PopulatinMU -0.707  0.463  0.500  0.500                            
## PopulatinRE -0.655  0.429  0.463  0.463  0.463                     
## PopulatinUM -0.678  0.444  0.479  0.479  0.479  0.444              
## PopulatinVA -0.707  0.463  0.500  0.500  0.500  0.463  0.479       
## PopulatinYE -0.707  0.463  0.500  0.500  0.500  0.463  0.479  0.500
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see ?isSingular

Grath Lab

anova(Via_lmer_Grath)

## Type III Analysis of Variance Table with Satterthwaite's method
##             Sum Sq  Mean Sq NumDF  DenDF F value Pr(>F)
## Population 0.07035 0.035175     2 27.293  2.1926 0.1309

summary(Via_lmer_Grath)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ Population + (1 | Line:Population)
##    Data: d_Via[d_Via$Supervisor.PI == "Grath", ]
## 
## REML criterion at convergence: -155.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -3.14211 -0.56745  0.04161  0.55412  2.60033 
## 
## Random effects:
##  Groups          Name        Variance Std.Dev.
##  Line:Population (Intercept) 0.003394 0.05826 
##  Residual                    0.016043 0.12666 
## Number of obs: 147, groups:  Line:Population, 30
## 
## Fixed effects:
##              Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)   0.69027    0.02612 28.41694  26.424   <2e-16 ***
## PopulationMU -0.07424    0.03664 27.57805  -2.026   0.0525 .  
## PopulationRE -0.05444    0.03664 27.57805  -1.486   0.1487    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) PpltMU
## PopulatinMU -0.713       
## PopulatinRE -0.713  0.508

Gibert Lab

anova(Via_lmer_Gibert)

## Type III Analysis of Variance Table with Satterthwaite's method
##            Sum Sq  Mean Sq NumDF  DenDF F value    Pr(>F)    
## Population 0.6029 0.075362     8 154.72  8.1983 3.224e-09 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

summary(Via_lmer_Gibert)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ProportionEggtoAdultSurvival ~ Population + (1 | Line:Population) +  
##     (1 | Batch)
##    Data: d_Via[d_Via$Supervisor.PI == "Gibert", ]
## 
## REML criterion at convergence: -687.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.8895 -0.5220  0.0038  0.5549  2.4861 
## 
## Random effects:
##  Groups          Name        Variance  Std.Dev.
##  Line:Population (Intercept) 1.064e-02 0.103146
##  Batch           (Intercept) 6.195e-05 0.007871
##  Residual                    9.192e-03 0.095877
## Number of obs: 532, groups:  Line:Population, 169; Batch, 3
## 
## Fixed effects:
##               Estimate Std. Error        df t value Pr(>|t|)    
## (Intercept)    0.75693    0.02658  51.26227  28.476  < 2e-16 ***
## PopulationGI  -0.14773    0.03986 155.28178  -3.706 0.000292 ***
## PopulationKA  -0.03422    0.03696 155.95970  -0.926 0.355981    
## PopulationMA  -0.11278    0.03691 154.48413  -3.056 0.002647 ** 
## PopulationMU  -0.03629    0.03692 155.28351  -0.983 0.327104    
## PopulationRE  -0.11002    0.03921 155.98273  -2.806 0.005655 ** 
## PopulationUM  -0.04241    0.03792 155.09608  -1.118 0.265136    
## PopulationVA  -0.12079    0.03692 155.28351  -3.272 0.001316 ** 
## PopulationYE  -0.24601    0.03692 155.28351  -6.664 4.38e-10 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) PpltGI PpltKA PpltMA PpltMU PpltRE PpltUM PpltVA
## PopulatinGI -0.640                                                 
## PopulatinKA -0.691  0.462                                          
## PopulatinMA -0.693  0.463  0.500                                   
## PopulatinMU -0.692  0.463  0.500  0.500                            
## PopulatinRE -0.651  0.437  0.470  0.470  0.471                     
## PopulatinUM -0.673  0.451  0.487  0.487  0.487  0.458              
## PopulatinVA -0.692  0.463  0.500  0.500  0.500  0.471  0.487       
## PopulatinYE -0.692  0.463  0.500  0.500  0.500  0.471  0.487  0.500

Lab correlations

Developmental time

Egg-to-pupa developmental time

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Schmidt Lab

Egg-to-adult developmental time

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Gibert Lab

Grath Lab

Hoedjes Lab

Schmidt Lab

Stamenkovic-Radak Lab

Zwaan Lab

Lab correlations

Dry weight

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Colinet Lab

Hoedjes Lab

Onder Lab

Lab correlations

Thorax length

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Kozeretska Lab

Posnien Lab

Ritchie Lab

Schmidt Lab

Lab correlations

Wing area - Right

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Onder Lab

Posnien Lab

Ritchie Lab

Stamenkovic-Radak Lab

Lab correlations

Wing area - Left

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Onder Lab

Posnien Lab

Ritchie Lab

Stamenkovic-Radak Lab

Lab correlations

Fecundity

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Billeter Lab

Fricke Lab

Lab correlations

Lifespan

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Flatt Lab

Parsch Lab

Pasyukova Lab

Survival Analyses

Flatt Lab

Parsch Lab

Pasyukova Lab

Lab correlations

Cold-shock mortality

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Gonzalez Lab

Kozeretska Lab

Vieira Lab

Lab correlations

Chill-coma recovery time

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Vieira Lab

Mensh Lab

Lab correlations

Heat-shock mortality

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Parsch Lab

Vieira Lab

Lab correlations

Diapause

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Bergland Lab

Flatt Lab

Schlotterer Lab

Lab correlations

Circadian eclosion timing

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Wegener Lab

Meeting w Eran and Chris

Eran to make circular figures for both CET and LA
no sexual dimorphism for period. YAY!
auto detection of phase is not reliable in the software - Eran wants to drop that trait in his data.
Watson Ftest
Circular data ANOVA : there are R packages
more cold adapted was expected to eclose later (to wait for the warmer period of the day) but it is not the case
LD : log2 transformation
Activity : any transformation
Period : ?transformation
Eran’s L:D -> D:D (circ)
Chris sent littoralis paper - can use our diapause results to compare!

DrosEU Phenotyping Working Group Updates

by Esra Durmaz Mitchell

Last updated: 2022-01-20

Phenotyping effort

Samples

Phenotyping experiments

Overview

Methods

Phenotyping

Viability

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Developmental time

Egg-to-pupa developmental time

Schmidt Lab

Egg-to-adult developmental time

Gibert, Grath, Hoedjes, Schmidt, Stamenkovic-Radak and Zwaan Labs

Dry weight

Colinet, Hoedjes and Onder Labs

Thorax length

Kozeretska, Posnien, Ritchie and Schmidt Labs

Wing area

Onder, Posnien, Ritchie and Stamenkovic-Radak Labs

Fecundity

Billeter and Fricke Labs

Lifespan

Flatt, Parsch and Pasyukova Labs

Cold-shock mortality

Gonzalez, Kozeretska and Vieira Labs

Chill-coma recovery time

Vieira and Mensh Labs

Heat-shock mortality

Parsch and Vieira Labs

Diapause

Bergland, Flatt and Schlotterer Labs

Circadian eclosion timing

Wegener Lab

Locomotor activity

Tauber Lab

Starvation resistance

Gonzalez, Onder and Pasyukova Labs

Pigmentation

Abbott, Gibert and Schmidt

Data reformatting

Data check

Results I : Linear Models

Viability

Data reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Schmidt Lab

Stamenkovic-Radak Lab

Zwaan Lab

Hoedjes Lab

Grath Lab

Gibert Lab

Lab correlations

Developmental time

Egg-to-pupa developmental time

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Schmidt Lab

Egg-to-adult developmental time

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Gibert Lab

Grath Lab

Hoedjes Lab

Schmidt Lab

Stamenkovic-Radak Lab

Zwaan Lab

Lab correlations

Dry weight

Data Reformatting

Descriptive Statistics

Plots and Linear Models by Lab

Colinet Lab

Hoedjes Lab

Onder Lab